【正文】
幾個方面:(1)測試集是否真正的開放域,即覆蓋的范圍是不是盡量的寬廣;(2)測試集的提問方式能否反應(yīng)用戶實際使用時的情況;(3)測試指標(biāo)能否有效、合理的比較各個問答系統(tǒng)的性能。n 構(gòu)建更為合理的打分標(biāo)準(zhǔn)目前的評分標(biāo)準(zhǔn)只是從問答系統(tǒng)返回的答案的角度進(jìn)行打分,此外,如果還考慮問答系統(tǒng)返回答案的文檔,打分會更合理。而對于其他類型的問題,如程序型提問、解釋型提問、摘要型提問、比較型提問等等,應(yīng)該有一個更客觀的打分標(biāo)準(zhǔn)。n 逐步擴(kuò)大用戶提問的廣度和深度我們希望能與國內(nèi)外問答檢索領(lǐng)域的團(tuán)隊合作,在各個研究小組的共同參與下,互相驗證彼此的研究成果,完善以漢語為主的QA測試集,合成權(quán)威的相關(guān)結(jié)果集,一起推動漢語問答檢索技術(shù)研究與應(yīng)用。參考文獻(xiàn):[1] Ellen M. Voorhees, Dawn M. Tice. The TREC8 Question Answering Track Evaluation[A]. The Eighth Text REtrieval Conference (TREC8), Spec Pub 500246, Washington DC: NIST, 1999, 7782.[2] Ellen M. Voorhees. Overview of the TREC 2003 question answering track[A]. In Proceedings of the Twelfth Text REtrieval Conference (TREC 2003), 2003.[3] Ellen M. Voorhees. Overview of the TREC9 Question Answering Track[A]. The Ninth Text REtrieval Conference (TREC9), Spec Pub 500249, Washington DC: NIST, 2000, 7782.[4] Ellen M. Voorhees. Overview of the TREC2001 Question Answering Track[A]. The Tenth Text REtrieval Conference (TREC01), Spec Pub 500250, Washington DC: NIST, 2001, 4251.[5] Ellen M. Voorhees. Overview of the TREC2002 Question Answering Track[A]. The Eleventh Text REtrieval Conference (TREC02), Spec Pub 500251,Washington DC: NIST, 2002.[6] John Burger et al. 2001. Issues, Tasks and Program Structures to Roadmap Research in Question amp。 Answering (Qamp。A) [A]. [7] Junichi Fukumoto, Tsuneaki Kato and Fumito Masui. Question Answering Challenge (QAC1): An Evaluation of QA Tasks at the NTCIR Workshop 3[A]. In Proc. of AAAI Spring Symposium: New Directions in Question Answering, , 2003. [8] Xiaoyan Li, W. Bruce Croft, Evaluating QuestionAnswering Techniques in Chinese[A]. Computer Science Department University of Massachusetts, Amherst, MA , 2001. [9] B. Magnini, S. Romagnoli, A. Vallin, J. Herrera, A. Pe241。as, V. Peinado, F. Verdejo, M. de Rijke. Creating the DISEQuA Corpus: a Test Set for Multilingual Question Answering[A]. Working Notes for the CLEF 2003 Workshop, 2122 August, Trondheim, Norway, 2003.[10] B. Magnini, S. Romagnoli, A. Vallin, J. Herrera, A. Pe241。as, V. Peinado, F. Verdejo, M. de Rijke. The Multiple Language Question Answering Track at CLEF 2003[A]. Working Notes for the CLEF 2003 Workshop, 2122 August, Trondheim, Norway, 2003.[11] John Burger, Claire Cardie, Vinay Chaudhri, et al. Issues, Tasks and Program Structures to Roadmap Research in Question amp。 Answering (Qamp。A). October 2000. 9